The overall goal of this research program is to utilize electrophysiological analyses techniques, specifically EG spectral analyses, slow wave sleep modeling, and methods from chaos theory to explore the neural activity which underlies disturbances in sleep and waking in recovering alcoholics and young men and women at risk for alcohol dependence. Remarkably few studies have evaluated sleep in alcoholism, especially over an extended period of recovery. This scarcity of sleep studies in alcoholism stands in sharp contrast to the frequency of sleep-related complaints in patients with alcoholism. Our previous studies as well as those of others suggest that specific EEG measures may be particularly valuable in: 1) predicting relapse in recovering alcoholics 2) exploring individual differences related to genetic vulnerability to alcoholism and 3) quantifying the effects of chronic ethanol exposure. In addition, we have developed predictive models of the relationship between specific changes in REM and NREM sleep and neurohormonal measures. The present study represents a new collaboration between the PI's. Dr. Gillin presently has pilot funding to collect sleep data in alcoholics with and without alcohol-induced mood-disorder with depressive features, hospitalized during early recovery, and 3 months following admission. The present study will allow such data to be analyzed using new theoretical and analytical methodology. An ongoing collaboration with Dr. Schuckit will provide access to at risk subjects. The study has 5 specific aims: 1) To utilize multivariate measures of the sleep and waking EEG to compare controls to alcoholics with and without alcohol- induce mood-disorder with depressive features, during early recover 2) To further explore our preliminary finding that certain REM sleep changes observed in alcoholics during early recovery are predictive of relapse, 3) To begin to evaluate and compare the sleep patterns of normal young adults with and without a family history of alcoholism 4) To evaluate the secretion of growth hormone and cortisol during the night in relation to EEG sleep 5) To develop a predictive model of sleep changes based on the Borbely 2 process model of sleep regulation and measures of neuroendocrine function. Within the center, this component, Dr. Ehlers, Dr. Gillin, and Dr. Schuckit provide an important role in linking the basic science components act to provide grounding to this clinical component. As new paradigms/drugs become available this information can be incorporated into potential new clinical pilot project.